SaaS Benchmarks
Stage-banded benchmarks for MRR growth, NRR, CAC payback, magic number, burn multiple, gross margin, and quick ratio. Sourced from Donum customer aggregates and public market data.
How to read this
Three bands per metric: below-median (bottom 50% of the stage cohort), median (middle band), and top-quartile (top 25%). "Elite" refers to top-decile performance often seen in category-leading companies near IPO.
MRR growth rate (MoM)
- Pre-seed to seed: median 15-25%, top quartile 30-50%+
- Seed to Series A: median 10-15%, top quartile 20-30%
- Series A to B: median 6-10%, top quartile 12-18%
- Series B+: median 3-6%, top quartile 8-12%
Net Revenue Retention (NRR)
- Below median: <100%
- Median: 105-110%
- Top quartile: 115-125%
- Elite: 130%+
Gross Revenue Retention (GRR)
- Below median: <85%
- Median: 88-92%
- Top quartile: 93-96%
- Elite: 97%+
CAC Payback Period
- Below median: 24+ months
- Median: 15-24 months
- Top quartile: 9-15 months
- Elite: <9 months
Magic Number
- Pull back: <0.5
- Steady: 0.5-0.75
- Invest more: 0.75-1
- Elite: >1.5
Burn Multiple (Sacks)
- Amazing: <1
- Great: 1-1.5
- Good: 1.5-2
- Suspect: 2-3
- Bad: >3
Gross Margin
- Below SaaS floor: <65%
- Acceptable: 65-75%
- Healthy: 75-82%
- Elite: 83%+
Quick Ratio (SaaS)
- Below median: <2
- Median: 2-4
- Top quartile: 4-8
- Elite: >8
Rule of 40
- Failing: <20
- Marginal: 20-40
- Healthy: 40-60
- Elite: 60+
Caveats
Benchmarks are reference points, not targets. A company consistently below median on one dimension can be elite overall. The right question is not "are we at the median" but "is our position on this metric consistent with the go-to-market motion we've chosen".
Frequently asked questions
Where do these benchmarks come from?
Two sources: (1) anonymised aggregates from Donum customers who opted into benchmarking, (2) publicly reported metrics from SEC filings and investor materials for public SaaS companies. We do not include individual customer data.
Why are the ranges so wide?
Because stage, category, and go-to-market motion create real variance. A 50-person infra company at $5M ARR looks nothing like a 50-person prosumer app at $5M ARR. The bands are the realistic spread, not the ideal.